Mobile, tablet, voice: each one changed how your customer reaches you. None changed who was on the other end. That is why this time is different.
When mobile arrived, companies rebuilt screens, flows and navigation. It was real work, and it was all on the presentation layer.
Behind every new surface there was still a person. Someone who noticed the page disagreed with the email, called to check, and complained when nothing worked. The interface changed. The loop around it did not.
That loop was doing four jobs nobody wrote down. It corrected ambiguity locally, reconciled channels that disagreed, stayed inside the intended path, and reported when something broke. Mobile inherited all four for free. So did tablet. So did voice.
An agent inherits none of them. It takes whatever the first channel it reads says, completes whatever the system allows, and leaves without a word.
This is the first user that changes the epistemic relationship between a company and whoever consumes it, not just the interface. Until now a company learned about its own defects because users told it. With this user, the signal that used to arrive by default does not arrive.
So the mobile playbook answers a different question. New screens, responsive layouts and a team to own the channel all assume feedback keeps flowing and the only open problem is presentation. Here the open problem is what happens when it stops.
None of this depends on agents buying in volume. Today they mostly search and compare. It depends on what the company stops hearing while they do.
Before you file the next kind of user under “mobile again”, ask one thing: does it still repair, reconcile and report? If it does, you are redesigning a surface. If it doesn’t, you have lost a sensor, and no amount of surface work gets it back.